(Circulation. 2000;101:264.)
© 2000 American Heart Association, Inc.
Clinical Investigation and Reports |
From Framingham Heart Study, National Heart, Lung, and Blood Institute, Framingham, Mass (K.A.P., M.G.L., J.C.E., P.A.S., P.W.F.W., D.L.); Institute for the Prevention of Cardiovascular Disease, Cardiovascular Division, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Mass (K.A.P., G.H.T., I.L., M.A.M., J.E.M.); National Heart, Lung and Blood Institute, Bethesda, Md (P.W.F.W., D.L.); and Department of Mathematics, Boston University, Boston, Mass (M.G.L., J.C.E., P.A.S., R.B.D.).
Correspondence to Daniel Levy, MD, Framingham Heart Study, Five Thurber Street, Framingham, MA 01702.
| Abstract |
|---|
|
|
|---|
Methods and ResultsWe studied 1193 men and 1459 women after the exclusion of subjects with known cardiovascular disease and those receiving anticoagulant or antihypertensive therapy. Linear regression models were used to evaluate SBP and DBP as predictors of fibrinolytic and hemostatic factor levels in separate sex models, with adjustment for age, body mass index, smoking, diabetes, total cholesterol, HDL, triglycerides, alcohol intake, and estrogen use (in women). In both sexes, levels of plasminogen activator inhibitor and tissue plasminogen activator antigen were positively related to SBP and DBP (P<0.001). Plasma viscosity was positively related to SBP (P=0.008) and DBP (P=0.001) in women only. There was no association between SBP or DBP and fibrinogen, factor VII, or von Willebrand factor in either sex.
ConclusionsThese data suggest that impaired fibrinolysis may play an important role in the pathogenesis of cardiovascular disease in hypertensive patients.
Key Words: hypertension fibrinolysis cardiovascular diseases
| Introduction |
|---|
|
|
|---|
| Methods |
|---|
|
|
|---|
Data were collected from each subject during a visit to the Framingham
Heart Study clinic. Subjects were studied throughout the year, with no
predetermined preference for season of the year. Blood pressures were
measured to the nearest 2 mm Hg with a mercury column
sphygmomanometer on the left arm after the subject had been seated
quietly for 5 minutes. Two readings obtained by the physician were
averaged to calculate the SBP and DBP. The use of any medications,
including antihypertensives, anticoagulants, and hormone replacement
therapy (HRT), was determined at the clinic visit. Subjects were asked
to stop taking aspirin products 1 week before the clinic visit
unless medically mandatory. No subjects were excluded for aspirin use.
Current smokers were defined as individuals who smoked on average at
least 1 cigarette per day during the year before the examination.
Alcohol consumption was assessed on the basis of the number of drinks
reported per week of beer (12 oz), wine (4 oz), and liquor (1 oz). Body
mass index (BMI) was computed by dividing the weight (in kilograms) by
the square of the height (in meters). Subjects were classified as
diabetic if they were taking an oral hypoglycemic agent or insulin at
the time of the examination, if they had a prior history of diabetes,
or if their fasting glucose level was
140 mg/dL.
Subjects were excluded for the following reasons: (1) history of CVD (n=304), (2) current use of antihypertensive medication (n= 591), (3) current use of oral anticoagulants (n=6), (4) missing covariates (n=63), or (5) missing laboratory data on hemostatic or fibrinolytic factors, except for viscosity (n=183). Prevalent CVD included coronary disease, stroke, intermittent claudication, or congestive heart failure. Standard Framingham definitions were used for these events on the basis of a review by a panel of 3 physicians.38
Blood Sampling and Analysis
Blood samples were collected between 8 and 9 AM from
an antecubital vein with subjects in the supine position after an
overnight fast to minimize circadian variations. For the determination
of plasma levels of PAI-1 and tPA antigen, blood was anticoagulated
with 3.8% trisodium citrate (9:1, vol/vol) and kept on crushed ice
until centrifugation. Plasma was separated by
centrifugation at 2500g for 30 minutes at
4°C. Plasma aliquots were quickly frozen and stored at -70°C for
subsequent analysis. PAI-1 antigen levels were determined with
a commercially available sandwich enzyme-linked immunosorbent assay
(ELISA) according to the description of Declerck et
al39 (TintElize PAI-1; Biopool AB). Levels of tPA
antigen also were obtained with the use of an ELISA according to a
procedure described by Ranby et al.40 Fibrinogen levels
were determined according to the method of Clauss.41
Factor VII antigen levels were measured with a commercially available
ELISA kit (Diagnostica Stago). vWF antigen levels were
measured according to an ELISA technique.42 Plasma
viscosity was measured in specimens previously frozen at -80°C with
the use of a Brookfield viscometer. The intra-assay coefficient of
variation in our laboratory was 9.6% for PAI-1, 5.5% for TPA, 2.6%
for fibrinogen, 8.8% for vWF, and 3.0% for FVII.
Statistical Analysis
Data on clinical characteristics are presented as
mean±SD. Logarithmic transformation was performed for variables
that were severely skewed (PAI-1, tPA antigen). Linear regression
models were used to evaluate hemostatic factors in relation to blood
pressure as a continuous variable.43 44 Unadjusted
results are presented first, followed by the primary
analyses, with adjustment for age, BMI, lipids (total
cholesterol, HDL cholesterol,
triglycerides), smoking, diabetes, alcohol intake, and HRT
(in women). Data on clinical characteristics are presented as
mean±SD. The results of regression models are presented as the
increment per 10 mm Hg of blood pressure, with 95% CI values.
Statistical significance was established at a 2-sided
level of
0.05.
| Results |
|---|
|
|
|---|
140 mm Hg
systolic or 90 mm Hg diastolic, in accordance
with the sixth report of the Joint National Committee on Detection,
Evaluation and Treatment of High Blood Pressure.45 Of the
496 subjects defined as hypertensive, 354 (71%) were hypertensive at 1
or more prior examinations, with an average duration of 11 years from
the examination at which hypertension was first documented.
|
Associations With SBP
In unadjusted analyses, SBP was associated with levels of
all 6 hemostatic factors in both sexes (in men, P<0.0001
for log PAI-1, log tPA, and fibrinogen, and 0.01< P<0.025
for FVII, vWF, and plasma viscosity; in women, P<0.0001 for
all 6 factors). Results from regression models that adjusted for
covariates (age, BMI, smoking, diabetes, alcohol intake,
triglycerides, total and HDL cholesterol, and
HRT in women) are shown in Table 2
. Increased
levels of PAI-1 antigen and tPA antigen remained highly associated with
increased SBP (P<0.001 in both sexes). In addition, plasma
viscosity was directly associated with SBP in women
(P=0.008) but not in men (P=0.91), whereas
neither fibrinogen, FVII, nor vWF remained associated with SBP in men
(P>0.15) or in women (P>0.25).
|
Figure 1
shows PAI-1 antigen levels (mean
and 95% CI) for men and women as a function of SBP adjusted for
covariates. Men had higher PAI-1 antigen levels than women; moreover,
PAI-1 levels increased markedly with increased SBP in both men and
women. Similar results are shown for tPA antigen in Figure 2
. In men, the full regression model
accounted for 33% of the variance for log PAI-1 and 30% of the
variance for log tPA (SBP accounted for 1.5% and 0.9%, respectively).
Corresponding data for women were 39% of the variance for log PAI-1
and 38% of the variance for log tPA (with SBP accounting for 0.5% and
0.7%).
|
|
Associations With DBP
In unadjusted analyses, DBP was associated with levels of
log PAI-1 and log tPA antigen in both sexes (P<0.0001),
with FVII (in men P=0.003; in women P<0.0001),
and with fibrinogen and plasma viscosity in women
(P<0.0001). Results from regression models with covariates
are shown in Table 3
. Increased levels of
PAI-1 and tPA antigen remained highly associated with increased DBP
(P<0.001 in both sexes). Plasma viscosity remained
associated with DBP in women (P=0.001), and vWF was
negatively related to DBP in men (P=0.03). Neither
fibrinogen, FVII, nor plasma viscosity was associated with DBP in men
(P>0.13), nor were fibrinogen, FVII, or vWF in women
(P>0.75). In men, the regression model accounted for 33%
of the variance for log PAI-1 and 31% of the variance for log tPA (DBP
accounted for 1.9% and 1.6%, respectively). Corresponding data for
women were 39% of the variance for log PAI-1 and 38% of the variance
for log tPA (with DBP accounting for 0.5% and 0.6%).
|
Partial correlation analyses showed that in men, the unadjusted associations of SBP with fibrinogen and vWF were accounted for by their joint relations with age, whereas the associations of SBP and DBP with FVII were accounted for by age and BMI. The relation of plasma viscosity with SBP was accounted for by age. Partial correlation analyses in women demonstrated that age accounted for the association of SBP with vWF, whereas the relations of SBP and DBP with fibrinogen and FVII were accounted for by age and BMI.
| Discussion |
|---|
|
|
|---|
Prior studies that evaluated the association of hemostatic factors and hypertension are limited by smaller sample size16 34 35 36 or lack of adjustment for 1 or more of the following variables: metabolic factors (diabetes, triglycerides),28 29 alcohol intake,28 29 32 HRT,13 17 48 or use of antihypertensive therapy.32 33 34 All of these are potential confounders and may alter hemostatic levels.
A study by Jannson et al31 compared 84 untreated hypertensive subjects who had elevated cholesterol levels with 55 control subjects matched for age, sex, and BMI. They found a positive association of plasma PAI-1 levels with triglycerides, total cholesterol, and DBP. Although these results support the presence of a decreased fibrinolytic state in hypertension, several differences among the 2 groups were not adjusted for in the analysis. The hypertensive group had higher triglyceride levels, a higher proportion of smokers, and a tendency toward a higher BMI, all of which may decrease fibrinolysis. More recently, Wall et al35 studied 39 young male subjects with untreated borderline hypertension and 17 normotensive control subjects matched for BMI, smoking, alcohol consumption, and total cholesterol, LDL, HDL, and triglyceride levels. The 24-hour ambulatory blood pressure recordings were performed, and an average value was computed for each subject. Borderline hypertensive subjects had a higher concentration of tPA antigen than did control subjects, whereas there were no significant differences in tPA activity or PAI-1 antigen between groups. In a stepwise regression analysis, BMI was the strongest predictor of tPA activity and PAI-1 antigen, whereas 24-hour mean arterial blood pressure emerged as the most powerful predictor of tPA antigen level. The results of the study by Wall et al35 support the presence of decreased fibrinolytic potential in a very selective group of young men with borderline hypertension who are unlikely to have significant undiagnosed arteriosclerosis.
In contrast to these small case-control studies, as part of a large cohort study, the Northern Sweden Monitoring of Trends and Determinants in Cardiovascular Disease (MONICA) study, Eliasson et al32 performed a cross-sectional analysis of 1558 participants, of whom each had blood samples drawn to determine plasma fibrinogen, tPA, and PAI-1 activity. In men, but not in women, DBP predicted low tPA activity and high PAI-1 activity. PAI-1 activity increased with triglycerides and BMI in both sexes. A high PAI-1 activity was correlated with SBP in men and women. Of note, subjects with a history of myocardial infarction or stroke, with diabetes, hypertensive subjects receiving treatment, and women receiving HRT were included in the analysis. A substudy addressed this issue by excluding all of these subjects, leaving a total of 1260 participants; in the multiple linear regression model, the association of tPA activity with DBP was no longer significant. BMI and hypertriglyceridemia were the strongest predictors of PAI-1 and tPA activity in that study.32
Hypertension has been closely related to obesity, hyperlipidemia, and glucose intolerance, all elements of the "insulin resistance syndrome."6 7 Furthermore, plasma insulin is thought to be an important regulator of PAI-1 activity, either directly or indirectly through plasma triglyceride concentration.49 In our study, PAI-1 antigen levels remained associated with increasing blood pressure after adjustment for known confounders, including hypertriglyceridemia and diabetes. Therefore, although PAI-1 activity is strongly correlated with hyperinsulinemia and triglycerides, it appears that PAI-1 antigen levels are independently associated with hypertension. The mechanism whereby increasing blood pressure may result in impaired fibrinolysis is unclear, but it may be related to the increase in shear stress or endothelial dysfunction.
Other Hemostatic Factors and Hypertension
Because fibrinogen, plasma viscosity, vWF, and FVII levels are all
associated with the risk of CVD, we also determined their associations
with hypertension. After adjustment for potential confounders, plasma
viscosity was associated with DBP but not SBP. There was no significant
association of fibrinogen or FVII with blood pressure and a modest
inverse association of vWF antigen with DBP in men.
The association of plasma viscosity with hypertension has been shown by several investigators16 17 18 19 20 to be independent of hematocrit.16 18 As a result, plasma proteins, and in particular fibrinogen, an important determinant of plasma viscosity, are thought to contribute to increased plasma viscosity. In contrast to our findings, the Edinburgh Artery Study found a relation of plasma viscosity to SBP and DBP in men, which persisted independent of hematocrit and fibrinogen.18 20 Elevated fibrinogen levels have not been consistently associated with hypertension.16 32 In the MONICA study, fibrinogen was correlated with blood pressure in both men and women; however, in a regression analysis, fibrinogen remained independently associated with SBP in women only. Other abnormalities of hemostatic factors, such as elevated levels of vWF33 and FVII,34 have also been associated with hypertension, but possible confounding variables were not adjusted for in the analysis, including smoking, serum glucose, BMI, and hypertension treatment. The adjustment for these variables in our study may explain the lack of an independent association of blood pressure with fibrinogen, FVII, and vWF antigen. This is supported by our partial correlation analyses, which showed that differences in age and BMI accounted for the association of these variables with blood pressure in our study sample.
Study Strengths and Limitations
In comparison with previous studies that examined the relations of
hemostatic and fibrinolytic factors to blood pressure, our study has
the benefit of a large sample and a statistical model that better
addresses known confounders, including triglycerides and
diabetes, in the analysis of PAI-1. In addition, the Framingham
Heart Study has the benefit of a broad population, which in this case
was selected with respect to treatment with antihypertensive
medication. A cross-sectional study such as ours, however, is always
limited by unrecognized confounding variables. In addition, the use
of mild criteria for hypertension as defined by the Sixth Report of the
Joint National Committee on Detection, Evaluation and Treatment of High
Blood Pressure45 and shorter duration of hypertension may
account for the negative associations found with fibrinogen and
vWF.
Conclusions
Hypertension is known to be a major risk factor for CVD. Although
antihypertensive treatment does reduce the risk of CVD, the results are
usually not commensurate to what would be predicted on the basis of
observational studies.4 This discrepancy may be in part
attributable to contributing factors such as a prothrombic state, which
is related to blood pressure level but is not corrected with
antihypertensive treatment. Our finding that increased blood pressure
is associated with decreased fibrinolytic potential and increased
plasma viscosity supports the presence of a prothrombic state in
hypertension. These data suggest that impaired
fibrinolysis may play an important role in the
pathogenesis of CVD in hypertensive subjects. Modification of this
hypercoagulable state may have a beneficial effect on the increased CVD
risk in subjects with hypertension.
| Acknowledgments |
|---|
Received July 13, 1998; revision received August 26, 1999; accepted September 15, 1999.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
C. Giannarelli, A. Virdis, F. De Negri, E. Duranti, A. Magagna, L. Ghiadoni, A. Salvetti, and S. Taddei Tissue-Type Plasminogen Activator Release in Healthy Subjects and Hypertensive Patients: Relationship With {beta}-Adrenergic Receptors and the Nitric Oxide Pathway Hypertension, August 1, 2008; 52(2): 314 - 321. [Abstract] [Full Text] [PDF] |
||||
![]() |
W. Lieb, M. J. Pencina, T. J. Wang, M. G. Larson, K. J. Lanier, E. J. Benjamin, D. Levy, G. H. Tofler, J. B. Meigs, C. Newton-Cheh, et al. Association of Parental Hypertension With Concentrations of Select Biomarkers in Nonhypertensive Offspring Hypertension, August 1, 2008; 52(2): 381 - 386. [Abstract] [Full Text] [PDF] |
||||
![]() |
V. F. Panoulas, G. S. Metsios, A. V. Pace, H. John, G. J. Treharne, M. J. Banks, and G. D. Kitas Hypertension in rheumatoid arthritis Rheumatology, May 29, 2008; (2008) ken159v2. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. J. Wang, P. Gona, M. G. Larson, D. Levy, E. J. Benjamin, G. H. Tofler, P. F. Jacques, J. B. Meigs, N. Rifai, J. Selhub, et al. Multiple Biomarkers and the Risk of Incident Hypertension Hypertension, March 1, 2007; 49(3): 432 - 438. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Shankar, J. J. Wang, E. Rochtchina, and P. Mitchell Positive Association Between Plasma Fibrinogen Level and Incident Hypertension Among Men: Population-Based Cohort Study Hypertension, December 1, 2006; 48(6): 1043 - 1049. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Kathiresan, S. B. Gabriel, Q. Yang, A. L. Lochner, M. G. Larson, D. Levy, G. H. Tofler, J. N. Hirschhorn, and C. J. O'Donnell Comprehensive Survey of Common Genetic Variation at the Plasminogen Activator Inhibitor-1 Locus and Relations to Circulating Plasminogen Activator Inhibitor-1 Levels Circulation, September 20, 2005; 112(12): 1728 - 1735. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. Arikan and S. Sen Endothelial Damage and Hemostatic Markers in Patients with Uncomplicated Mild-to-Moderate Hypertension and Relationship with Risk Factors Clinical and Applied Thrombosis/Hemostasis, April 1, 2005; 11(2): 147 - 159. [Abstract] [PDF] |
||||
![]() |
R. Rossi, E. Chiurlia, A. Nuzzo, E. Cioni, G. Origliani, and M. G. Modena Flow-mediated vasodilation and the risk of developing hypertension in healthy postmenopausal women J. Am. Coll. Cardiol., October 19, 2004; 44(8): 1636 - 1640. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Hayashi, E. J. Boyko, D. L. Leonetti, M. J. McNeely, L. Newell-Morris, S. E. Kahn, and W. Y. Fujimoto Visceral Adiposity Is an Independent Predictor of Incident Hypertension in Japanese Americans Ann Intern Med, June 15, 2004; 140(12): 992 - 1000. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. D. Sesso, J. E. Buring, N. Rifai, G. J. Blake, J. M. Gaziano, and P. M. Ridker C-Reactive Protein and the Risk of Developing Hypertension JAMA, December 10, 2003; 290(22): 2945 - 2951. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Hayashi, E. J. Boyko, D. L. Leonetti, M. J. McNeely, L. Newell-Morris, S. E. Kahn, and W. Y. Fujimoto Visceral Adiposity and the Prevalence of Hypertension in Japanese Americans Circulation, October 7, 2003; 108(14): 1718 - 1723. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. Hognert, B. Ghanoum, H. Gustafsson, I. Milsom, and K. Manhem Acute Effects of Transdermal 17{beta}-Estradiol on Hemostatic Variables after 24-hour Treatment Clinical and Applied Thrombosis/Hemostasis, July 1, 2002; 8(3): 239 - 243. [Abstract] [PDF] |
||||
![]() |
C. G.C. Spencer, D. Gurney, A. D. Blann, D. G. Beevers, and G. Y.H. Lip Von Willebrand Factor, Soluble P-Selectin, and Target Organ Damage in Hypertension: A Substudy of the Anglo-Scandinavian Cardiac Outcomes Trial (ASCOT) Hypertension, July 1, 2002; 40(1): 61 - 66. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. GLICK Screening for traditional risk factors for cardiovascular disease: A review for oral health care providers J Am Dent Assoc, March 1, 2002; 133(3): 291 - 300. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. U. Chae, R. T. Lee, N. Rifai, and P. M. Ridker Blood Pressure and Inflammation in Apparently Healthy Men Hypertension, September 1, 2001; 38(3): 399 - 403. [Abstract] [Full Text] [PDF] |
||||
![]() |
N. J. Brown, L. J. Murphey, N. Srikuma, N. Koschachuhanan, G. H. Williams, and D. E. Vaughan Interactive Effect of PAI-1 4G/5G Genotype and Salt Intake on PAI-1 Antigen Arterioscler. Thromb. Vasc. Biol., June 1, 2001; 21(6): 1071 - 1077. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. Y. H. Lip Target Organ Damage and the Prothrombotic State in Hypertension Hypertension, December 1, 2000; 36(6): 975 - 977. [Full Text] [PDF] |
||||
![]() |
L. A. Sechi, L. Zingaro, C. Catena, D. Casaccio, and S. De Marchi Relationship of Fibrinogen Levels and Hemostatic Abnormalities With Organ Damage in Hypertension Hypertension, December 1, 2000; 36(6): 978 - 985. [Abstract] [Full Text] [PDF] |
||||
![]() |
H.-C. Chen, J. L. Bouchie, A. S. Perez, A. C. Clermont, S. Izumo, J. Hampe, and E. P. Feener Role of the Angiotensin AT1 Receptor in Rat Aortic and Cardiac PAI-1 Gene Expression Arterioscler. Thromb. Vasc. Biol., October 1, 2000; 20(10): 2297 - 2302. [Abstract] [Full Text] [PDF] |
||||
![]() |
Hypertension Associated with Prothrombotic State Journal Watch (General), January 28, 2000; 2000(128): 8 - 8. [Full Text] |
||||
![]() |
G. Y. H. Lip and A. D. Blann Does Hypertension Confer a Prothrombotic State? : Virchow’s Triad Revisited Circulation, January 25, 2000; 101(3): 218 - 220. [Full Text] [PDF] |
||||
| |||||||||||||||||||||||